Attaining effective genetically encoded optical voltage-indicators has been a longstanding goal in neuroscience research and is a key near-term aim of the BRAIN Initiative. Unlike small molecule sensors or hybrids of fluorescent proteins with organic molecules, optical voltage-indicators that can be fully encoded genetically are readily combined with genetic tools and viral delivery methods that enable long-term expression and chronic imaging studies without addition of exogenous agents. Genetically encoded Ca2+-sensors offer similar targeting advantages, but Ca2+-imaging fails to reveal individual spikes in many neuron types, poorly captures sub- threshold membrane dynamics, and has insufficient temporal resolution to capture spike timing to better than ~50-100 ms. Voltage-indicators directly sense the membrane potential and promise faithful reporting of spike waveforms, spike bursts and sub-threshold dynamics, in cells targeted by their genetic class or connectivity. An ideal voltage-indicator would produce large fluorescence responses, to facilitate spike detection, and have millisecond-scale kinetics, to study synchrony and spike-timing aspects of neural coding. However, prior protein voltage-indicators have generally suffered performance-limiting tradeoffs between modest brightness, sluggish kinetics, and limited signaling dynamic range in response to action potentials. To date, no protein voltage-indicator combines the attributes needed for accurate reporting of voltage activity in behaving animals. However, if such a sensor emerged, this would likely have even greater impact on brain science than the surge in research enabled by recent advanced versions of the GCaMP Ca2+-indicator. This proposal seeks to create broad voltage-imaging capabilities and involves two Co-PDs who are highly experienced in fluorescence imaging of neural activity. Working collaboratively, we recently created two new classes of voltage-indicators, of distinct colors and voltage-sensing mechanisms, each of which has substantially superior signaling fidelity than earlier protein voltage-indicators while offering faster kinetics and higher brightness. Using thes two sensor types, we have imaged fast spike trains in cultured neurons and brain slices. Calculations using signal detection theory show our indicators are now on the brink of transitioning into a mainstay approach to monitor large numbers of individual neurons in behaving animals. To enact this, we will use novel massively parallel methods to screen variants of our protein indicators at 100-1000 greater throughput than screening methods used previously in the field. We will validate and iteratively optimize the resulting indicators in cultred neurons, mammalian brain slices, and behaving flies, nematodes and mice, by using signal detection theory to benchmark indicator performance. To accompany these voltage-indicators, we will also create imaging instrumentation custom-designed for high-speed (~1 kHz) voltage-imaging in awake head-restrained and freely behaving mice. If our work succeeds, it will be a game-changer for brain research, propelling studies of how cells and circuits function normally and go awry in disease.